METHOD FOR PREDICTING POSSIBILITY OF BRAIN AMYLOID BETA ACCUMULATION

FIELD OF THE INVENTION

The present invention relates to a method of predicting the likelihood of brain amyloid-beta deposition in order to predict the onset of diseases caused by brain amyloid-beta deposition, including Alzheimer's disease, in Koreans, and more specifically, to a method of providing information predicting the likelihood of brain amyloid-beta deposition, and a composition and a kit for predicting the likelihood of brain amyloid-beta deposition.

BACKGROUND OF THE INVENTION

Alzheimer's disease (AD) is a representative degenerative brain disease that accounts for about 70% of dementia, and is the most common cause of dementia in the elderly. Genetic factors play an important role in the pathogenesis of Alzheimer's disease (AD) because heritability is estimated to be 58% to 79%, which indicates a genetic predisposition. Apolipoprotein E (APOE) 24 is a well-known risk variant for AD, and recent genome-wide association studies (GWAS) have discovered a number of genetic risk variants for AD. However, a large proportion of AD heritability is still unexplained.

Meanwhile, early changes in the brains of AD patients are deposition of amyloid-beta (Aβ) in the brain, followed by tau deposition, neurodegeneration, and cognitive impairment. That is, amyloid-beta begins to accumulate in the brain 10 to 15 years before clinical symptoms of AD appear. Therefore, detecting or discovering individuals with brain Aβ deposition is of utmost importance for the early diagnosis and early treatment of AD. Conventionally, deposition of amyloid-beta in the brain was examined by positron emission tomography (PET) or cerebrospinal fluid analysis. However, PET is mainly performed in large hospitals due to high imaging costs and limited access, and causes risk factors such as radiation exposure, and cerebrospinal fluid analysis has disadvantages in that an invasive lumbar puncture is required and reliability varies depending on organs. Therefore, there is a need for biomarkers capable of predicting amyloid-beta deposition in the brain in a low-cost and non-invasive manner.

Since genomes differ greatly between races, genetic factors associated with Alzheimer's disease and amyloid-beta deposition in the brain differ between races. Most of previous studies on Alzheimer's disease genes have been conducted on Europeans, and genetic variants studies on non-Europeans are insufficient. In addition, studies on brain amyloid-beta-related genes, early diagnostic markers for Alzheimer's disease, are very rare even in foreign countries (Yan Q, et al. Mol Psychiatry. 2018:1-13; Apostoloval L G et al. JAMA Neurol. 2018: 328-341; Raghavan N S, et al. JAMA Neurol. 2020:1288-1298) and have not been conducted in Korea.

Therefore, in order to predict brain amyloid-beta deposition, which is an early diagnostic marker for Alzheimer's disease in Koreans, by a low-cost and minimally invasive method (basic clinical information and genetic information), a predictive model based on Korean data is required. In addition, verification procedures should be performed to ensure that the same model works even in an independent cohort.

SUMMARY OF THE INVENTION

An object of the present invention is to provide a method of providing information for predicting the likelihood of brain amyloid-beta deposition based on a combination of individual characteristics and genetic information.

Another object of the present invention is to provide a composition for predicting the likelihood of brain amyloid-beta deposition, which may be used to identify genetic information.

Still another object of the present invention is to provide a kit for predicting the likelihood of brain amyloid-beta deposition, the kit comprising the composition.

The present inventors have conducted GWAS analysis to identify single nucleotide polymorphisms (SNPs) associated with brain amyloid-beta deposition using a large sample of the Korean population. As a result, the present inventors have identified six novel SNPs for amyloid-beta deposition and demonstrated their associations in an independent cohort of the Korean population. In addition, the present inventors have developed a novel amyloid-beta prediction model by combining these novel SNP genotypes with individual characteristics and APOE genotypes, thereby completing the present invention that provides a method of predicting the likelihood of brain amyloid-beta deposition in the Korean population and a related composition.

One aspect of the present invention provides a method of providing information for predicting the likelihood of brain amyloid-beta deposition through a combination of demographic characteristics, APOE ε4 genotypes and SNP genotypes. More specifically, the method of providing information for predicting the likelihood of brain amyloid-beta deposition comprises steps of: a) collecting individual characteristics including sex, age, and education level from an individual; b) measuring an APOE ε4 genotype and at least one SNP genotype selected from the group consisting of rs73375428, rs6978259, rs2903923, rs3828947, rs6958464, and rs11983537, in a biological sample from the individual; and c) predicting the likelihood of brain amyloid-beta deposition by combining the collected individual characteristics and the measured APOE ε4 genotype and SNP genotype.

As used herein, “likelihood of brain amyloid-beta deposition” means the risk of brain amyloid-beta deposition in an individual, and it can be said that the higher the risk, the higher the likelihood of developing a disease caused by amyloid-beta deposition. Here, the disease caused by amyloid-beta deposition may be Alzheimer's disease, Lewy body dementia, cerebral amyloid angiopathy, etc., without being limited thereto. This prediction of the likelihood of brain amyloid-beta deposition is distinct from techniques for diagnosing Alzheimer's disease or Alzheimer's dementia in an individual. In the present invention, a case where amyloid-beta is likely to be deposited in the brain or has already been deposited in the brain may be determined as positive.

Steps a) to c) will now be described in detail

Step a) is a process of collecting individual characteristics from an individual that needs to be tested for brain amyloid-beta deposition.

The individual is a subject in whom the likelihood of amyloid-beta deposition in the cerebrum is to be predicted, and may be a patient having a family history of a disease caused by brain amyloid-beta deposition (e.g., Alzheimer's disease, Lewy body dementia, or cerebral amyloid angiopathy) or suspected of having a brain disease due to some clinical symptoms due to memory loss, language impairment, and judgment decline, or may be a person with no specific clinical symptoms.

The individual characteristics are basic characteristics of an individual, such as demographic characteristics, and may include sex, age, and education level.

The sex refers to a biological sex, and it is known that women are more likely to develop Alzheimer's disease than men. The age refers to the actual age calculated at the time of examination with the actual time of birth as the starting point, and it is known that the likelihood of amyloid-beta deposition increases with age, and especially, Alzheimer's disease mainly occurs in the elderly aged 65 years or older. The education level means the entire level of education including formal education courses, and in the case of dropouts, the level of education was calculated excluding the year in which the student dropped out. Recently, it has been reported that the risk of dementia increases in the case of low education. Therefore, the sex, age, and education level of an individual may be significant variables in predicting the likelihood of brain amyloid-beta deposition.

Step b) is a process of measuring the level of a biomarker present in a biological sample.

The biological sample refers to a sample collected from an individual and usable for predicting the likelihood of brain amyloid-beta deposition, and may be, for example, blood, cerebrospinal fluid, tears, saliva, urine, or the like. In the present invention, it is preferable to use blood, saliva, oral mucosa, or hair, which is relatively painless during the collection process and inexpensive, compared to cerebrospinal fluid.

According to one embodiment of the present invention, the biological sample may be one selected from the group consisting of whole blood, plasma, saliva, oral mucosa, and hair.

The biomarker is a substance that is generally detectable in a biological sample, and examples thereof include all organic biomolecules such as polypeptides, proteins, nucleic acids, genes, lipids, glycolipids, glycoproteins, and sugars, which can indicate biological changes. The present invention is characterized in that the APOE ε4 genotype and the SNP genotype are used as biomarkers.

The APOE gene is classified into three alleles, called ε2 (cys112, cys158), ε3 (cys112, arg158), and ε4 (arg112, arg158), and everyone inherits one APOE allele from each parent. Thus, the genotype of an individual is determined as ε2/ε2 , ε2/ε3 , ε2/ε4 , ε3/ε3 , ε3/ε4, or ε4/ε4. Among them, the ε4 allele is found in about 20% of the population and is known to increase the risk of developing Alzheimer's dementia, and thus APOE ε4 is a risk factor highly associated with the development of AD. Therefore, possession of the APOE ε4 genotype can be a significant variable in predicting the likelihood of brain amyloid-beta deposition, and the risk of brain amyloid-beta deposition can increase depending on the number of APOE ε4 alleles.

According to an embodiment of the present invention, when the likelihood of brain amyloid-beta deposition in Koreans is predicted by combining the APOE ε4 genotype with individual characteristics for Koreans, the Area Under Curve (AUC) value can be improved.

Here, AUC means the total area under the ROC curve calculated through ROC analysis, and it is known that the AUC value is 0.5 for a random model and converges to 1 as the predictive performance of the model improves.

The SNP refers to the existence of two or more alleles at one locus, and means that only a single base in the polymorphic site differs. In the present invention, the SNP genotype may be at least one selected from the group consisting of rs73375428, rs6978259, rs2903923, rs3828947, rs6958464 and rs11983537, which were confirmed to be associated with brain amyloid-beta deposition through GWAS analysis performed on the Korean population. Such rs73375428, rs6978259, rs2903923, rs3828947, rs6958464 and rs11983537 are located in the intron of the coiled-coil domain containing 146 (CCDC146) gene in human chromosome 7, show high linkage disequilibrium (LD: r²>0.75) with each other, and are located at locus 76907550, locus 76909167, locus 76907750, locus 76908199, locus 76909035, and locus 76908690 of chromosome 7, respectively. These six SNPs have not previously been reported for their association with brain amyloid-beta deposition, and their relationship with brain amyloid-beta deposition in Koreans has also not been reported.

In addition, according to one embodiment of the present invention, AUC can be improved when the likelihood of brain amyloid-beta deposition is predicted by combining individual characteristics and the SNP genotype together with the APOE ε4 genotype.

Step c) is a process of preparing a brain amyloid-beta deposition prediction model by combining individual characteristics, APOE ε4 genotype and SNP genotype.

According to one embodiment of the present invention, step c) may be performed by setting individual characteristics, APOE ε4 genotype, and SNP genotype as independent variables, setting specific coefficients for them as dependent variables, and preparing a brain amyloid-beta deposition prediction model through statistical analysis.

According to one embodiment of the present invention, the statistical analysis may be performed using an analysis method selected from the group consisting of linear or nonlinear regression analysis method, linear or nonlinear classification analysis method, ANOVA, neural network analysis method, deep neural network analysis method, genetic analysis method, support vector machine analysis method, hierarchical analysis or clustering analysis method, hierarchical algorithm or Kernel principal components analysis method using a decision tree, Markov Blanket analysis method, recursive feature elimination or entropy-based recursive feature elimination analysis method, forward floating search or backward floating search analysis method, and combinations thereof.

According to an embodiment of the present invention, a brain amyloid-beta deposition prediction model was prepared by analyzing the relationship between individual characteristics, APOE ε4 genotype and SNP genotype, set as independent variables, and each specific coefficient set as dependent variables, by multivariate logistic regression analysis.

According to one embodiment of the present invention, the brain amyloid-beta deposition prediction model may be represented by Equation 1 below.

Aβ deposition=−0.90+(0.0026*age)+(0.164*sex)+(0.0177*education level)+(1.653*APOE ε4 genotype)+(α*SNP genotype) [Equation 1]

(In the Equation 1, α is −0.612 when SNP is rs6978259 , −0.652 when SNP is rs6958464, −0.658 when SNP is rs11983537, −0.571 when SNP is rs3828947, −0.612 when SNP is rs2903923, and −0.633 when SNP is rs73375428.)

According to one embodiment of the present invention, when the likelihood of brain amyloid-beta deposition is predicted by combining individual characteristics, APOE ε4 genotype, and SNP genotype, AUC can be improved. In this case, the AUC in the brain amyloid-beta deposition prediction model is 0.74 or more regardless of the type of SNP (rs6978259, 0.748; rs6958464, 0.749; rs11983537, 0.749; rs3828947, 0.751; rs2903923, 0.748; and rs73375428, 0.749), suggesting that the accuracy is higher than that when only individual characteristics are used alone (0.506) or that when individual characteristics and APOE ε4 genotype are used in combination (0.723).

It can be said that the higher the value derived from the brain amyloid-beta deposition prediction model, the higher the likelihood of brain amyloid-beta deposition in the subject to be tested.

In addition, another aspect of the present invention provides a composition for predicting the likelihood of brain amyloid-beta deposition, the composition containing an agent capable of detecting APOE ε4 genotype and at least one SNP genotype selected from the group consisting of rs73375428, rs6978259, rs2903923, rs3828947, rs6958464, and rs11983537.

In the present invention, the expression “composition for predicting the likelihood of brain amyloid-beta deposition” refers to a composition for a series of actions of predicting brain amyloid-beta deposition by detecting nucleotide sequences including APOE ε4 genotype and SNP genotype. Using this composition, the risk of developing a disease by brain amyloid-beta deposition or whether a disease has been developed by brain amyloid-beta deposition may be diagnosed at an early stage.

Here, the APOE ε4 genotype and the six SNP genotypes are the same as described above.

The agent is capable of detecting biomarkers, that is, the APOE ε4 genotype and the SNP genotype, for predicting the likelihood of brain amyloid-beta deposition.

According to one embodiment of the present invention, the agent may be at least one selected from the group consisting of a primer set, a probe, and an antisense oligonucleotide, which specifically bind to a nucleotide sequence including APOE ε4 genotype or SNP genotype. In the present invention, the agent is preferably a primer set designed to specifically amplify a specific region of a gene based on nucleic acid sequence information of genes, or a probe.

The term “specifically bind” means that the binding affinity to a target substance is superior to the binding affinity to other substances to the extent that the presence or absence of the target substance can be detected by binding.

The primer set refers to a combination of primers that are short-stranded RNA or DNA sequences that recognize all or part of a nucleotide sequence including the APOE ε4 genotype or SNP genotype. For example, the primer set may be a primer set of any combination of forward and reverse primers capable of binding to a nucleotide sequence including the APOE ε4 genotype or SNP genotype. Since the nucleic acid sequence of the primer is a sequence that does not match the non-target sequence present in the sample, it may amplify only the target gene sequence containing the complementary primer binding site, and may give high specificity when the primer does not cause non-specific amplification.

The probe refers to a substance capable of specifically binding to a target substance to be detected in a sample, and capable of specifically detecting the presence of the target substance in the sample through the binding. The constituent material of the probe is a material capable of specifically binding to the target substance, and examples thereof include, but are not limited to, peptide nucleic acid (PNA), locked nucleic acid (LNA), peptides, polypeptides, proteins, RNA, DNA, and the like.

Still another aspect of the present invention provides a kit for predicting the likelihood of brain amyloid-beta deposition, the kit comprising the composition.

In the present invention, the expression “kit for predicting the likelihood of brain amyloid-beta deposition” refers to a substance capable of predicting the likelihood of brain amyloid-beta deposition through a biological sample collected from a test subject. Using the kit, the risk of developing a disease by brain amyloid-beta deposition or whether a disease has been developed by brain amyloid-beta deposition may be diagnosed at an early stage in a rapid and convenient manner. Examples of the kit include, but are not limited to, a polymerase chain reaction (PCR) kit, a DNA chip kit, etc. In addition, the kit may further comprise one or more other compositions, solutions or devices suitable for assay.

According to one embodiment of the present invention, the kit may comprise essential elements necessary for performing polymerase chain reaction (PCR). In this case, the kit may comprise, in addition to a primer set that specifically binds to a nucleotide sequence including the APOE ε4 genotype or at least one SNP genotypes selected from the group consisting of rs73375428, rs6978259, rs2903923, rs3828947, rs6958464 and rs11983537, a test tube or other appropriate container, a reaction buffer (which may have various pHs and magnesium concentrations depending on conditions), deoxynucleotides (dNTPs), enzymes such as Taq polymerase and reverse transcriptase, DNase and RNAse inhibitors, DEPC-treated water, sterilized water, etc.

According to one embodiment of the present invention, the kit may comprise essential elements necessary for performing DNA chip assay. In this case, the kit may comprise a substrate to which a cDNA or oligonucleotide corresponding to a nucleotide sequence including the APOE ε4 genotype or at least one SNP genotype selected from the group consisting of rs73375428, rs6978259, rs2903923, rs3828947, rs6958464 and rs11983537 or a fragment of the nucleotide sequence is attached, and a reagent, an agent, an enzyme, and the like for producing a fluorescently labeled probe. In addition, the substrate may comprise a cDNA or oligonucleotide corresponding to a control gene or a fragment thereof.

The method of providing information for predicting the likelihood of brain amyloid-beta deposition according to the present invention may effectively screen individuals with a high likelihood of brain amyloid-beta deposition at an early stage, thereby providing an appropriate method of inhibiting the development of a disease caused by brain amyloid-beta deposition or treating an individual having the disease.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a quantile-quantile plot of P values according to an Example of the present invention.

FIG. 2 is a graph showing SNPs located on chromosomes 19 and 7 in the whole genome and significantly associated with amyloid-beta deposition, according to an Example of the present invention.

FIG. 3 is a graph showing the chromosomal location of rs6978259 according to an Example of the present invention.

FIG. 4 shows amyloid positron emission tomography (PET) images of subjects for voxel-based morphometry analysis according to an Example of the present invention. Specifically, (A) depicts images showing that Aβ deposition decreases in a subject with the rs6978259 allele, and (B) depicts images showing that Aβ deposition increases in a subject with the APOE ε4 allele.

FIG. 5 shows AUC values of datasets for development and validation of models for rs6978259 according to an Example of the present invention.

FIG. 6 shows AUC values of datasets for development and validation of models for rs6958464 according to an Example of the present invention.

FIG. 7 shows AUC values of datasets for development and validation of models for rs11983537 according to an Example of the present invention.

FIG. 8 shows AUC values of datasets for development and validation of models for rs3828947 according to an Example of the present invention.

FIG. 9 shows AUC values of datasets for development and validation of models for rs2903923 according to an Example of the present invention.

FIG. 10 shows AUC values of datasets for development and validation of models for rs73375428 according to an Example of the present invention.

DETAILED DESCRIPTION OF THE INVENTION

Hereinafter, the present invention will be described in more detail. However, these descriptions are provided for illustrative purposes only to help understand the present invention, and the scope of the present invention is not limited by these exemplary descriptions.

Example 1. Development of Model for Predicting Likelihood of Brain Amyloid-Beta Deposition

1-1. Recruitment of Participants

For a dataset development of a model, a total of 1,214 Koreans were recruited from 14 domestic hospitals from January 2013 to July 2019. Among them, 923 participants were recruited from the Samsung Medical Center, 201 participants were recruited from a multicenter study of the Korean Brain Aging Study for the Early Diagnosis and Prediction of AD (KBASE-V), and 90 participants were recruited from a multicenter study of Clinical Research Platform based on Dementia Cohort.

For a dataset for validation, data from 306 Korean participants were collected from the Biobank of the Chronic Cerebrovascular Disease Consortium, recruited from 2016 to 2018. This was part of the ongoing Biobank Innovation for chronic Cerebrovascular disease With ALZheimer's disease Study (BICWALZS) and the Center for Convergence Research of Neurological Disorders.

Among the recruited test subjects, the present inventors included participants (i) who were diagnosed with amnestic mild cognitive impairment (aMCI), AD dementia (ADD), or were cognitively unimpaired (CU) based on detailed neuropsychological tests, and (ii) who underwent amyloid PET imaging, and excluded participants who had a causative genetic mutation for AD, such as PSEN1, PSEN2, and APP. Quality control (QC) was performed as described later, and for data for model development, a total of 1,190 participants (383 with CU, 330 with aMCI, and 477 with ADD) were used, and for data for validation, a total of 284 participants (46 with CU, 167 with aMCI, and 71 with ADD) were used. All participants provided written informed consent, and the study was approved by the Institutional Review Board of each center.

As a result, the baseline demographic characteristics of the two datasets (data for model development and data for verification) are shown in Table 1 below. Participants for model development were younger and had a higher proportion of females than data for validation. As expected, in both datasets, amyloid-beta (Aβ) positive participants confirmed by PET had higher rates of aMCI and AD than Aβ negative participants. In addition, in the data for validation, Aβ-positive participants were older than Aβ-negative participants.

TABLE 1

Data for model development

Demographic
Total
Aβ negative
Aβ positive

characteristics
(n = 1,190)
(n = 561)
(n = 629)
P value*

Age, year (SD)
70.07
70.06
70.07
0.990

(8.75)
(8.16)
(9.25)

Female, n (%)
680
310
370
0.215

(57.1)
(55.3)
(58.8)

Education, year
11.02
10.89
11.13
0.390

(SD)
(4.86)
(5.04)
(4.70)

Diagnosis, n (%)

<0.001

CU
383
326
57

(32.2)
(58.1)
(9.1)

aMCI
330
172
158

(27.7)
(30.7)
(25.1)

ADD
477
63
414

(40.1)
(11.2)
(65.8)

Data for validation

Demographic
Total
Aβ negative
Aβ positive

characteristics
(n = 284)
(n = 180)
(n = 104)
P value*

Age, year (SD)
72.67
71.76
74.25
0.006

(7.32)
(7.31)
(7.10)

Female, n (%)
184
122
62
0.165

(64.8)
(67.8)
(59.6)

Education, year
8.34
7.77
9.11
0.050

(SD)
(5.21)
(5.19)
(5.15)

Diagnosis, n (%)

<0.001

CU
46
43
3

(16.2)
(23.9)
(2.9)

aMCI
167
125
42

(58.8)
(69.4)
(40.4)

ADD
71
12
59

(25.0)
(6.7)
(56.7)

*P value was calculated by comparing Aβ negative and Aβ positive participants.

P values calculated by comparing the data for model development and the data for validation were <0.001, 0.019, <0.001, and <0.001 for age, sex, education, and diagnosis, respectively.

Student's t test and chi-squared test were used for continuous and categorical variables, respectively.

1-2. Genotyping and Imputation

Participants were genotyped using the Illumina Asian Screening Array BeadChip (Illumina, USA) for data for model development and Affymetrix customized Korean chips (Affymetrix, CA, USA) for data for model validation, and SNP markers were analyzed. The present inventors conducted QC using PLINK software (version 1.9). Participants were excluded based on the following criteria: (i) call rate <95%, (ii) mismatch between reported and genetically inferred sex, (iii) deviation from each population parameter, (iv) excess heterozygosity rate (5 standard deviation from the mean), and (v) in cases of related pairs (identified with identity by descent (IBD) ≥0.125) within and between the datasets for model development and for model validation.

SNPs were excluded based on the following criteria: (i) call rate <98%, (ii) minor allele frequency (MAF) <1%, and (iii) a p value <1.0×10⁻⁶for the Hardy-Weinberg equilibrium test.

After QC, genome-wide imputation was performed using the Minimac4 software with all available reference haplotypes from HRC-r1.1 on the University of Michigan Imputation Server. For post-imputation QC, the present inventors excluded SNPs based on the following criteria: (i) poor imputation quality (R²≤0.8), and (ii) MAF ≤1%. Finally, 4,906,407 SNPs were analyzed.

1-3. Aβ PET Acquisition and Voxel-Based Morphometry Analysis

Amyloid-beta (Aβ) PET images were obtained using a Discovery STE PET/CT scanner (GE Medical Systems, USA). PET images were acquired for 20 min, starting at 90 min after intravenous injection of either ¹⁸F-florbetaben or ¹⁸F-flutemetamol. Aβ positivity or negativity was determined using visual assessments for FBB and FMM PET. A subset of participants in the cohort for model development (n=824) had Aβ PET data available for voxel-based morphometry (VBM) analysis.

For VBM analysis, the present inventors performed the following preprocessing using Statistical Parametric Mapping software version 12 (SPM, http://www.fil.ion/uc.ac.uk/spm) running on MATLAB (MathWorks 2014b): (1) co-registration of PET to T1-weighted structural MRI, (2) structural MRI segmentation and calculation of transformation matrix, (3) normalization of PET to a Montreal Neurological Institute (MNI) space, and (4) spatial smoothing with a Gaussian kernel of 8-mm full width at half maximum.

To calculate the standardized uptake value ratio (SUVR) for each PET image, the present inventors used two reference regions (the cerebellar cortex for FBB and pons for FMM). The masks of reference regions were obtained from the GAAIN website (http://www.GAAIN.org).

1-4. Statistical Analysis

{circle around (1)} GWAS Analysis

Logistic regression analysis was performed to determine the association between SNPs and Aβ positivity in a state in which age, gender, and population subgroup effects were adjusted, and the additive model, coded as 0, 1, and 2 according to the number of minor alleles, was used. In order to adjust for population subgroup effects, principal component analysis was performed on the genetic data, and the first, second, and third principal components were used as correction variables. Reported p values were two-tailed, and the present inventors defined a p value less than 5.0×10⁻⁸as being statistically significant and a p value less than 1.0×10⁻⁵as being statistically suggestive based on previous studies. Genomic inflation was evaluated by dividing the median of chi-squared statistics derived from GWAS analysis by the median (0.456) of a chi-square distribution with a degree of freedom of 1. For validation analysis, considering that the effect is expected to occur in the same direction as the result of the dataset for model development, one-tailed P values were reported. To check if SNPs were associated with Aβ positivity independent of APOE genotype, the present inventors performed conditional analysis by further adjusting for APOE genotype.

Furthermore, the present inventors performed P-value based meta-analysis and calculated the summary effect size by applying weights reflecting the standard errors from the study specific effect sizes.

As a result, referring to FIG. 1, a quantile-quantile (Q-Q) plot of p values revealed no genomic inflation (λ=1.008). Referring to FIG. 2, in the data for model development, 61 genome-wide significant SNPs on chromosome 19 (P value <5×10⁻⁸) were identified (see Table 2 below). However, all significant SNPs fell within the 500-kb region surrounding APOE gene and lost genome-wide significance when the APOE ε4 allele was adjusted. Outside of the APOE region, 38 SNPs on chromosomes 1, 7, 8, 12, and 22 showed genome-wide suggestive significance (P value <1.0×10⁻⁵) (see Table 3 below). In validation analysis, among 38 suggestive SNPs, 6 SNPs (rs73375428, rs6978259, rs2903923, rs3828947, rs6958464 and rs11983537) on chromosome 7 showed a significant association in the same direction as the results obtained from the data for model development. Two SNPs (rs6958464 and rs11983537) among the six SNPs were directly genotyped and the remaining SNPs (rs73375428, rs6978259, rs2903923 and rs3828947) were analyzed by imputation. In particular, two of the six SNPs (rs73375428 and rs2903923) showed genome-wide significant associations in the meta-analysis of the datasets for model development and validation (see Tables 4 and 5 below). When the effect of the APOE ε4 allele was adjusted, all of the six SNPs showed P values at genome-wide suggestive levels, and the associations were verified in the data for validation. As shown in Table 6 below, the identified six SNPs showed high linkage disequilibrium (LD: r²>0.75) with each other, and thus rs6978259 out of the six SNPs was used as a representative for subsequent analysis.

TABLE 2

Base pairs
Effective

SNP
Chromosome
(bp)
allele
OR
P value
P value*

rs8112196
19
45325309
A
0.632
9.12 × 10⁻⁸
0.086

rs6509172
19
45325391
T
0.616
7.06 × 10⁻⁹
0.044

rs1871045
19
45326768
T
0.613
6.32 × 10⁻⁹
0.056

rs73050216
19
45367502
C
0.527
1.93 × 10⁻¹⁴
0.049

rs41290098
19
45370278
T
0.547
4.59 × 10⁻¹³
0.165

rs12610605
19
45370838
G
1.917
1.21 × 10⁻¹⁴
0.354

rs404935
19
45372794
A
3.472
1.07 × 10⁻²⁹
0.193

rs395908
19
45373565
A
3.332
2.85 × 10⁻²⁸
0.451

rs34278513
19
45378144
T
2.868
1.12 × 10⁻²¹
0.515

rs412776
19
45379516
A
3.601
5.02 × 10⁻³¹
0.149

rs3865427
19
45380961
A
3.122
5.34 × 10⁻²⁴
0.314

rs6859
19
45382034
A
2.110
1.39 × 10⁻¹⁷
0.654

rs3852860
19
45382966
C
2.875
5.43 × 10⁻²⁷
0.113

rs3852861
19
45383061
G
2.850
1.19 × 10⁻²⁶
0.135

rs71352237
19
45383079
C
3.288
9.49 × 10⁻²⁵
0.493

rs34224078
19
45383115
G
3.288
9.49 × 10⁻²⁵
0.493

rs35879138
19
45383139
A
3.307
6.24 × 10⁻²⁵
0.430

rs166907
19
45386855
G
6.140
3.09 × 10⁻⁸
0.231

rs12972156
19
45387459
G
3.830
7.66 × 10⁻²⁹
0.427

rs12972970
19
45387596
A
3.830
7.66 × 10⁻²⁹
0.427

rs34342646
19
45388130
A
3.830
7.66 × 10⁻²⁹
0.427

rs283815
19
45390333
G
3.338
6.85 × 10⁻³²
0.298

rs6857
19
45392254
T
3.681
5.38 × 10⁻²⁸
0.261

rs71352238
19
45394336
C
3.748
2.03 × 10⁻²⁸
0.368

rs184017
19
45394969
G
3.289
2.24 × 10⁻³¹
0.396

rs157580
19
45395266
G
0.462
2.12 × 10⁻¹⁹
0.154

rs2075650
19
45395619
G
3.748
2.03 × 10⁻²⁸
0.368

rs157581
19
45395714
C
2.820
3.31 × 10⁻²⁷
0.966

rs34404554
19
45395909
G
3.716
3.22 × 10⁻²⁸
0.293

rs11556505
19
45396144
T
3.716
3.22 × 10⁻²⁸
0.293

rs157582
19
45396219
T
3.326
8.93 × 10⁻³²
0.373

rs59007384
19
45396665
T
2.802
3.38 × 10⁻²³
0.965

rs157583
19
45396673
T
6.677
5.65 × 10⁻⁹
0.117

rs157587
19
45398206
G
6.571
7.71 × 10⁻⁹
0.122

rs205909
19
45400775
G
6.677
5.65 × 10⁻⁹
0.117

rs491153
19
45402368
T
6.677
5.65 × 10⁻⁹
0.117

rs490243
19
45402470
T
6.677
5.65 × 10⁻⁹
0.117

rs417357
19
45403119
T
6.775
4.24 × 10⁻⁹
0.114

rs394819
19
45404579
T
6.775
4.24 × 10⁻⁹
0.114

rs405697
19
45404691
G
2.299
1.86 × 10⁻²¹
0.170

rs10119
19
45406673
A
4.948
1.21 × 10⁻⁴⁰
0.967

rs435380
19
45407118
A
6.576
7.63 × 10⁻⁹
0.133

rs446037
19
45407437
T
6.577
7.59 × 10⁻⁹
0.147

rs434132
19
45407720
G
6.577
7.59 × 10⁻⁹
0.147

rs439382
19
45408475
G
6.573
7.68 × 10⁻⁰
0.107

rs440446
19
45409167
G
2.318
3.06 × 10⁻²²
0.318

rs769449
19
45410002
A
4.601
4.83 × 10⁻³³
0.223

rs429358
19
45411941
C
5.275
9.03 × 10⁻⁴²
0.601

rs75627662
19
45413576
T
3.392
1.55 × 10⁻²⁸
0.790

rs439401
19
45414451
C
2.295
5.58 × 10⁻²²
0.279

rs10414043
19
45415713
A
4.285
8.20 × 10⁻³²
0.216

rs7256200
19
45415935
T
4.285
8.20 × 10⁻³²
0.216

rs483082
19
45416178
T
4.024
1.01 × 10⁻³⁷
0.213

rs584007
19
45416478
G
2.284
8.19 × 10⁻²²
0.310

rs438811
19
45416741
T
4.024
1.01 × 10⁻³⁷
0.213

rs5117
19
45418790
C
3.316
1.17 × 10⁻²⁷
0.921

rs12721046
19
45421254
A
4.497
6.23 × 10⁻³⁵
0.209

rs12721051
19
45422160
G
4.493
5.81 × 10⁻³⁵
0.214

rs56131196
19
45422846
A
4.517
3.73 × 10⁻³⁵
0.291

rs4420638
19
45422946
G
4.517
3.73 × 10⁻³⁵
0.291

rs157595
19
45425460
G
2.197
4.31 × 10⁻²⁰
0.520

*P values are statistical values of logistic regression analysis adjusted for the effect of the APOE ε4 allele

TABLE 3

Data for model
Data for validation

Base pairs
Effective
development (n = 1190)
(n = 284)

SNP
Chromosome
(bp)
allele
OR
P value
OR
P value*

rs73375428
7
76907550
G
0.519
2.71 × 10⁻⁷
0.549
0.020

rs6978259
7
76909167
C
0.521
4.62 × 10⁻⁷
0.566
0.030

rs2903923
7
76907750
G
0.528
5.15 × 10⁻⁷
0.539
0.016

rs3828947
7
76908199
C
0.528
5.15 × 10⁻⁷
0.546
0.018

rs6958464
7
76909035
T
0.525
6.28 × 10⁻⁷
0.555
0.029

rs11983537
7
76908690
T
0.557
7.58 × 10⁻⁷
0.539
0.013

rs112599253
7
76931677
T
0.561
1.56 × 10⁻⁷
0.723
0.107

rs79761449
7
76925493
T
0.564
2.50 × 10⁻⁷
0.723
0.107

rs6971106
7
76929191
T
0.564
2.50 × 10⁻⁷
0.723
0.107

rs113931965
7
76930080
T
0.564
2.50 × 10⁻⁶
0.723
0.107

rs113014884
7
76930314
C
0.564
2.50 × 10⁻⁶
0.723
0.107

rs58731747
7
76935492
A
0.568
2.75 × 10⁻⁶
0.706
0.091

rs1077236
8
130640501
T
1.612
2.85 × 10⁻⁶
0.993
0.488

rs10092679
8
130682158
T
1.580
5.33 × 10⁻⁶
0.941
0.390

rs1196470
1
151706235
A
0.669
5.43 × 10⁻⁶
0.992
0.484

rs4872051
8
22945345
G
0.687
5.66 × 10⁻⁶
0.827
0.135

rs140005
22
36943544
A
1.488
5.76 × 10⁻⁶
1.035
0.424

rs140008
22
36944165
T
1.488
5.76 × 10⁻⁶
1.035
0.424

rs140010
22
36944356
C
1.488
5.76 × 10⁻⁶
1.035
0.424

rs68055908
8
130667115
T
1.579
5.85 × 10⁻⁶
0.976
0.455

rs59674745
7
76913192
C
0.511
5.86 × 10⁻⁶
0.831
0.308

rs6470745
8
130641921
G
1.576
6.36 × 10⁻⁶
1.002
0.497

rs6985032
8
130671748
T
1.561
7.46 × 10⁻⁶
0.958
0.418

rs4295627
8
130685457
G
1.569
7.61 × 10⁻⁶
0.910
0.333

rs28572791
8
130641782
C
1.569
7.92 × 10⁻⁶
1.033
0.436

rs2868782
7
76907862
A
0.517
8.06 × 10⁻⁶
0.781
0.259

rs7306151
12
110092612
C
0.619
8.24 × 10⁻⁶
0.903
0.316

rs6970348
7
76929400
A
0.610
8.27 × 10⁻⁶
0.780
0.153

rs140018
22
36946330
C
1.479
8.97 × 10⁻⁶
0.985
0.467

rs140019
22
36946491
G
1.479
8.97 × 10⁻⁶
0.985
0.467

rs140011
22
36945022
G
1.479
9.02 × 10⁻⁶
0.990
0.479

rs140012
22
36945289
C
1.479
9.02 × 10⁻⁶
0.990
0.479

rs140014
22
36945641
C
1.479
9.02 × 10⁻⁶
0.990
0.479

rs140016
22
36945847
C
1.479
9.02 × 10⁻⁶
0.990
0.479

rs140017
22
36945947
A
1.479
9.02 × 10⁻⁶
0.990
0.479

rs144786745
8
130658990
T
1.569
9.23 × 10⁻⁶
1.039
0.427

rs4766712
12
114491139
T
1.511
9.40 × 10⁻⁶
0.870
0.251

rs5022680
8
130685516
T
1.562
9.42 × 10⁻⁶
0.923
0.356

*P value is statistical values of the logistic regression analysis performed on date for validation.

TABLE 4

SNP
Chromosome: base pairs (bp)
Effective allele

rs73375428
7:76907550
G

rs6978259
7:76909167
C

rs2903923
7:76907750
G

rs3828947
7:76908199
C

rs6958464
7:76909035
T

rs11983537
7:76908690
T

TABLE 5

Data for model development
Data for validation
Meta-analysis

SNP
OR
P value
OR
P value
OR
P value

Analysis 1

rs73375428
0.519
2.71 × 10⁻⁷
0.550
0.020
0.526
3.35 × 10⁻⁸

rs6978259
0.522
4.62 × 10⁻⁷
0.566
0.030
0.521
8.54 × 10⁻⁸

rs2903923
0.529
5.15 × 10⁻⁷
0.539
0.016
0.536
4.97 × 10⁻⁸

rs3828947
0.529
5.15 × 10⁻⁷
0.547
0.018
0.536
5.59 × 10⁻⁸

rs6958464
0.526
6.28 × 10⁻⁷
0.555
0.028
0.525
4.06 × 10⁻⁷

rs11983537
0.558
7.58 × 10⁻⁷
0.539
0.013
0.563
5.92 × 10⁻⁸

Analysis 2

rs73375428
0.535
1.23 × 10⁻⁵
0.481
0.011
0.516
8.00 × 10⁻⁷

rs6978259
0.521
7.90 × 10⁻⁶
0.515
0.020
0.541
9.08 × 10⁻⁷

rs2903923
0.546
2.19 × 10⁻⁵
0.478
0.010
0.510
1.32 × 10⁻⁶

rs3828947
0.546
2.19 × 10⁻⁵
0.480
0.010
0.515
1.39 × 10⁻⁶

rs6958464
0.524
9.63 × 10⁻⁶
0.484
0.014
0.521
8.18 × 10⁻⁷

rs11983537
0.570
1.99 × 10⁻⁵
0.492
0.010
0.517
1.22 × 10⁻⁶

Analysis 1: Values of logistic regression analysis not adjusted for the effect of APOE ε4 allele.

Analysis 2: Values of logistic regression analysis adjusted for the effect of APOE ε4 allele.

TABLE 6

rs73375428
rs2903923
rs3828947
rs11983537
rs6958464
rs6978259

rs73375428
1
1.00
1.00
0.92
0.87
0.87

rs2903923
1.00
1
1.00
0.92
0.87
0.87

rs3828947
1.00
1.00
1
0.92
0.87
0.87

rs11983537
0.92
0.92
0.92
1
0.88
0.88

rs6958464
0.87
0.87
0.87
0.88
1
1.00

rs6978259
0.87
0.87
0.87
0.88
1.00
1

{circle around (2)} Effects of Newly Identified SNPs

After identifying associated SNPs, the present inventors calculated the risk of the six SNPs on Aβ deposition in all participants and at each cognitive level (CU, aMCI, and ADD).

Next, the present inventors performed voxel-wise PET image analysis to determine which regional Aβ deposition is associated with SNPs after adjusting for the effects of age, sex, genetic PCs, APOE genotype, and Aβ PET tracer type.

To test the clinical utility of the newly identified SNPs, the present inventors developed multivariable logistic models to predict Aβ positivity in each individual. To evaluate the performance of the logistic model, the present inventors measured the area under curve (AUC) from the receiver operating characteristic (ROC) curve analysis.

For internal validation, the present inventors conducted a 10-fold cross-validation with 100 repeats using the data for model development. The present inventors reported the mean AUC with 95% confidence interval (CI) of the model. As an external validation, parameters estimated from the data for model development were used to calculate the Aβ prediction performance in the data for validation. The present inventors used R software (http://www.rproject.org) and MATLAB for the statistical analyses and results visualization.

As a result, the characteristics of the participants for the number of minor alleles (C) of rs6978259 are shown in Table 7 below.

TABLE 7

C = 0
C = 1
C = 2

(n = 894)
(n = 278)
(n = 18)
P custom-character

*

Age, year (SD)
70.06
70.01
71.22
0.850

(8.74)
(8.86)
(7.62)

Education, year (SD)
11.21
10.43
10.38
0.054

(4.80)
(5.01)
(5.00)

Diagnosis, n (%)

0.014

CU
268
107
8

(30.0)
(38.5)
(44.4)

aMCI
247
76
7

(27.6)
(27.3)
(39.8)

ADD
379
95
3

(42.4)
(34.2)
(16.7)

Female, n (%)
495
173
12
0.093

(55.4)
(62.2)
(66.7)

APOE ε4, n (%)

0.012

Noncarrier
511
187
13

(57.2)
(67.3)
(72.2)

Heterozygote
306
65
4

(34.2)
(23.4)
(22.2)

Homozygote
77
26
1

(8.6)
(9.4)
(5.6)

Aβ positive, n (%)
507
120
2
<0.001

(56.7)
(43.2)
(11.1)

*P values were calculated using one-way ANOVA or Chi-square test, as appropriate.

As shown in Table 8 below, in the logistic model, the APOE ε4 allele was associated with a 5-fold higher risk of Aβ positivity (odds ratio [OR]=5.330; 95% CI=4.188-6.788), and rs6978259 was associated with a 2-fold lower 5 risk of Aβ positivity (OR=0.521. 95% CI=0.407-0.670). In the subgroup analysis, the association of rs6978259 with Aβ positivity was significant in the CU and aMCI groups but not in the ADD group, while the association of APOE ε4 was significant across all cognitive states.

TABLE 8

rs6978259
APOE ε4

OR (95% CI)
P value
OR (95% CI)
P value

Total
0.521 (0.407-0.670)
<0.001
5.330 (4.188-6.788)
<0.001

(n = 1190)

CU
0.503 (0.256-0.988)
0.046
3.885 (2.307-6.54)
<0.001

(n = 383)

aMCI
0.455 (0.281-0.735)
0.001
6.655 (4.101-10.8)
<0.001

(n = 330)

ADD
0.667 (0.365-1.219)
0.188
4.272 (2.428-7.516)
<0.001

(n = 477)

Referring to FIG. 4, in the VBM analysis, APOE ε4 was associated with increased Aβ deposition on the wide cortical areas of the frontal, parietal, and temporal lobes. It was confirmed that rs6978259 was associated with decreased Aβ deposition in the precuneus, lateral parietal, and medial frontal areas, independent of age, sex, education level, APOE ε4, and Aβ PET tracer type.

Several prediction models were developed to test the clinical utility of the APOE ε4 allele, and each model for the six SNPs (rs6978259, rs6958464, rs11983537, rs3828947, rs2903923 and rs73375428) was developed (see Equation 1). Referring to FIG. 5, in the 10-fold cross-validation with 100 repetitions, model 1 including only individual characteristics (age, sex, and level of education) showed an AUC of 0.506 (95% CI=0.500-0.512). Model 2 obtained by incorporating the APOE ε4 allele in model 1 showed increased prediction performance (AUC=0.723; 95% CI=01717-01729), and model 3 (Equation 1) obtained by incorporating rs6978259 in model 2 showed further increased prediction performance (AUC=0.748; 95% CI=0.742-0.755). When each model, trained in the data for model development, was tested in the data for validation, model 3 including individual characteristics, APOE ε4 genotype and rs6978259 showed a higher AUC value than model 1 and model 2 (model 1 AUC=0.509, model 2 AUC=0.693, and model 3 AUC=0.713). Referring to FIGS. 6 to 10, similar results were also observed for rs6958464, rs11983537, rs3828947, rs2903923 and rs73375428.

{circle around (3)} Analysis of Function

Finally, the present inventors characterized the function of the six identified SNPs by leveraging bioinformatic tools and previously reported results. First, the present inventors checked whether MAF of SNPs identified in this Example was similar to the MAF in the East Asian population using the 1000 Genomes Project dataset. To evaluate the genotype-specific expression of the identified SNPs in human brain tissues, the present inventors performed cis-expression quantitative trait loci (cis-eQTL) analysis through the Genotype-Tissue Expression portal (https://gtexportal.org).

As a result, as shown in FIG. 3, rs6978259 was found to be located in the intron of the coiled-coil domain containing 146 (CCDC146) gene.

These results suggest that there are differences in the types of SNPs that are highly associated with brain amyloid-beta deposition due to differences in cohorts, and that APOE ε4 genotype as a blood biomarker and at least one SNP genotype selected from the group consisting of rs73375428, rs6978259, rs2903923, rs3828947, rs6958464 and rs11983537 are combined and analyzed with individual characteristics including sex, age, and education level for Koreans, it is possible to effectively screen individuals with a high likelihood of brain amyloid-beta deposition at an early stage.

So far, the present disclosure has been described with reference to the embodiments. Those of ordinary skill in the art to which the present disclosure pertains will appreciate that the present disclosure may be embodied in modified forms without departing from the essential characteristics of the present disclosure. Therefore, the disclosed embodiments should be considered from an illustrative point of view, not from a restrictive point of view. The scope of the present disclosure is defined by the claims rather than the foregoing description, and all differences within the scope equivalent thereto should be construed as being included in the present disclosure.

METHOD FOR PREDICTING POSSIBILITY OF BRAIN AMYLOID BETA ACCUMULATION

Information

Publication Number

Date Filed

Date Published

Inventors

Original Assignees

CPC

International Classifications

Abstract

Description

Claims

Priority Claims (1)

PCT Information